I find that detailed questioning about a design that I made almost 40 years ago to be out of line. I still have that little amp, but I don't want to fire it up, just to satisfy some idle curiosity.
It is amazing that 2 separate measurements on the same page of an INDEPENDENT review, first can be so far apart in their implications, and one measurement, when read a certain way, can imply a synthetic inductance that is above and beyond what can be normally generated by an amplifier like the JC-1. Perhaps I should patent the concept.
It is amazing that 2 separate measurements on the same page of an INDEPENDENT review, first can be so far apart in their implications, and one measurement, when read a certain way, can imply a synthetic inductance that is above and beyond what can be normally generated by an amplifier like the JC-1. Perhaps I should patent the concept.
john curl said:
Sorry to have you scratching your head, John. Recall that the stated output impedance of the Bryston also under-estimated its output inductance. We simply don't know the details of how JA measured the output impedance at 20 kHz. We also don't know the details of the measurement of loaded frequency response, some of which, themselves, could possibly cause loaded frequency response curves that could possibly over-state the output inductance.
Whether or not the JC-1 could have an effective output inductance of 2 uH is something that is hard for me to speculate on, although frankly that does seem a bit high, whether it is due to wiring inductance or due to falling feedback factor against the HF open-loop output impedance. To be honest, I would think that your driving 60 MHz ft output devices with a fairly hot MOSFET source follower would yield a pretty low HF open loop output impedance out to beyond 100 kHz.
Cheers,
Bob
john curl said:
Mr Curl, I'm not trying to be rude you, but may I remind you that is was you who started talking about your 40 years old amp. So, don't be surprised that someone asked for more specific and relevant details. Calling this 'to satisfy some idle curiosity', that's, at least, 'out of line'.
And to all: Please, stop speculating, based on output impedance, about the value of the output coil. This is getting a never ending story. Just ask the manufacturer and pray he gives a reliable answer or take that (evil) coil out of the enclosure and measure the value by yourself.
Cheers,
john curl said:
If I were the designer, I would be flattered, honestly.
Unfortunately, I'm not blessed with your gifts, John.
So, why not be a little more easy going, life is so short. 🙂
/Hugo
john curl said:
Hi John,
I did some back of the envelope speculation on the JC-1, and I agree with you that it seems unlikely that it would have 2 uH of effective output inductance, at least in any ordinary way that we would normally first think of.
Let's say you are driving your output transistors with a MOSFET source follower biased at 100 mA. It will have a relatively resistive output impedance on the order of 2.5 ohms. Let's also say that you have 5 ohm base-stopper resistors on each of your nine output devices on each side. But they are effectively in parallel together, so the net effect of the base stopper resistors is to just add on the order of an additional half ohm. So the effective source impedance to the bases is about 3 ohms. Lets say that the beta of your output transistors is 50 flat out to 1 MHz (=> 50 MHz ft). The fact that you use a MOSFET source follower probably effectively shields your output bases from any high impedance at the VAS (I'm assuming this, but it needs to be verified).
So if we divide the source impedance of 3 ohms by the collective beta of 50, we get only about 0.06 ohms flat out to about 1 MHz, suggesting far less than 2 uH, even in the open loop output impedance with no reliance on NFB.
Apart from a measurement artifact, there may be one other way that the 2 uH could happen. If there is some unintentional coupling from output current back to input voltage, the 2 uH could indeed be synthesized. Again, this seems unlikely, as I believe the Stereophile measurements of loaded frequency response were done in balanced mode.
But it might not take much coupling like this at 50 kHz and above to create the appearance of 2 uH. And the unintended coupling could occur in the external test setup or inside your amplifier.
At the 2.83 volt drive into 2 ohms, there is about 1.4A rms flowing. Suppose that current flow somehow created an effective 10 mV at the input at 50 kHz, which would then be multiplied by the amp's closed loop gain of, say, 30. That would make 300 mV at the output which, against 1.4 amps, corresponds to 0.21 ohms of impedance. That would correspond to about 0.7 uH, closer to the ballpark, but not there yet. Also, 10 mV of unintentional coupling under those conditions at 50 kHz seems a bit much.
So, even though this scenario is plausible, it still seems a bit unlikely, and the mystery continues.
One final thought: If you haven't done so, measure the JC-1's output impedance with a signal generator and a spectrum analyzer. Set your generator to 10 V rms at 50 kHz and back-feed it into the JC-1 through a 1K resistor. The input of the JC-1 should be shorted. Place the spectrum analyzer directly across the JC-1's output terminals. If you get a 50 kHz component on the spectrum analyzer of 1 mV, the output impedance at that frequency is about 0.1 ohm.
Cheers,
Bob
Estuart, have you NOW realized that there is NO real output coil on the output of the JC-1 amp? That is the point of this thread!
By the way, who where you dating 40 years ago? And give me all the details! At least what books you were reading. Do you see how difficult it is to go back so far?
By the way, who where you dating 40 years ago? And give me all the details! At least what books you were reading. Do you see how difficult it is to go back so far?
estuart said:
Bob's original question referred to above (which I cannot find now) was what the output inductance of a typical tube amplifier would be because of the output transformer leakage reactance.
The better output transformers had leakage reactances of 4 - 8 mH. That was however included in the feedback loop, so that the true output equivalent series inductance would be lower. It would not have a simple relation to the feedback factor because the feedback loop would also include capacitave phase shift acting at h.f.
The figure quoted by estuart from information by Mr Stoet seems rather low but possible. I presume that Mr Stoet understood the question to refer to an amplifier's output, not just the OPT.
This sheds further sobering light on some of the very academic arguments raised here regarding possible audible effects of a mere few uH in series with the loudspeaker.
john curl said:
Hi John,
Bear in mind that no one on this thread, as far as I know, ever asserted that you had a real output coil in the JC-1.
I was dating a really hot one 40 years ago, and I remember it like it was yesterday. 1967 was a wonderful year!
Cheers,
Bob
Bob Cordell said:
Far out. I only turned 30 in January. You guys are a bunch of fossils.
Cheers,
Glen
Re: Re: Re: Re: Re: Re: Re: definition of 'rock solid'
Hi Edmond.
I don't think that there is anything wrong with my approach in this regard. I made an assumption based on my theoretical understanding of the amplifiers operation, and this has been confirmed on the test bench - which I think carries a bit more weight than in simulation.
Because, whith my test equiptment, the method I outlined allows me to quickly and accurately examine the open loop gain and phase response of the amplifier from 0.1Hz to 20MHz.
This gives a lot a valuable information, some of which can be difficult to measure with the loop closed. Also, it is not difficult to implement, especially so since my mainframe dual trace Tektroniks 551 oscilloscope has a Type O operational amplifier plug-in installed, which is eminently suitable for DC servo duties 🙂
If you're talking about the capacitor from the VAS to the inverting input, the same method works fine, but testing of the inner loop requires a bit of elaboration.
Cheers,
Glen
estuart said:
Hi Edmond.
I don't think that there is anything wrong with my approach in this regard. I made an assumption based on my theoretical understanding of the amplifiers operation, and this has been confirmed on the test bench - which I think carries a bit more weight than in simulation.
estuart said:
Because, whith my test equiptment, the method I outlined allows me to quickly and accurately examine the open loop gain and phase response of the amplifier from 0.1Hz to 20MHz.
This gives a lot a valuable information, some of which can be difficult to measure with the loop closed. Also, it is not difficult to implement, especially so since my mainframe dual trace Tektroniks 551 oscilloscope has a Type O operational amplifier plug-in installed, which is eminently suitable for DC servo duties 🙂
estuart said:
If you're talking about the capacitor from the VAS to the inverting input, the same method works fine, but testing of the inner loop requires a bit of elaboration.
Cheers,
Glen
Hi folks, this is your old geezer talking! 
Yes, back when I was thirty, I was dating both a 20 year old San Francisco stripper, who lived in North Beach, and a 20 year old Las Vegas showgirl, down on her luck at the moment.
How is it going for you?
Yes, back when I was thirty, I was dating both a 20 year old San Francisco stripper, who lived in North Beach, and a 20 year old Las Vegas showgirl, down on her luck at the moment. How is it going for you?
john curl said:Hi folks, this is your old geezer talking!
How is it going for you?
From San Francisco, hey? Did she have a moustache?
No, they were both real, beautiful young women. I have had lots of those. I hope you can find a little satisfaction in your life as well.
G.Kleinschmidt said:
So that put's it all in perspective.
Enjoy the present and check back once the testosterone settles down and experience tempers the the intellect.
Cheers, Mike.
MikeBettinger said:
🙄
I hope to expire before my intellect becomes as tempered as some..............
That's what we all say, Sonny! (When we are 30, and full of beans)
Bob, could you PLEASE check the SAME measurements of the Halcro DM-88 power amplifier, published in the August 2006 issue of 'Stereophile'.
Bob, could you PLEASE check the SAME measurements of the Halcro DM-88 power amplifier, published in the August 2006 issue of 'Stereophile'.
mikelm said:
The problem is when humans are treated like sophistcated machines which they aren't.
Not showing up on the lab tests doesn't mean it won't be audible, it just means we must ask different questions and find different tests.
Folks, I would like to say something important here.
I try to contribute my experience to this website. As a working audio design engineer for about 40 years , I have some experience to offer.
In my near retirement years, all I hope is to pass on some knowledge to those, who might follow. Why I have to put up with insults regarding my age and experience is beyond me.
I try to contribute my experience to this website. As a working audio design engineer for about 40 years , I have some experience to offer.
In my near retirement years, all I hope is to pass on some knowledge to those, who might follow. Why I have to put up with insults regarding my age and experience is beyond me.
- Status
- Not open for further replies.